0
$\begingroup$

I have asked some participants (29) to complete several tasks, and I have the results for the people that successfully completed it, and those who didn't. I have a breakdown by age (4 groups) and I am trying to prove ($H_0$) that age is independent for the successful completion of the task.

I am trying to understand whether I have to use chi square or ANOVA. I used anova and I was able to reject the $H_0$, but if I use Chi Square, I fail to reject the Nullhypothesis.

Can someone guide me on this, please?

$\endgroup$
2
  • $\begingroup$ Could you please clarify if your dependent variable is a count (number of completed tasks) or a binary feature (completed/not completed)? $\endgroup$ Aug 30, 2017 at 15:11
  • $\begingroup$ Hi, thanks a lot for your reply! It was a task that they had to do and they would either get a success or unsuccess. Therefore, it is a binary feature. Thanks again $\endgroup$ Aug 30, 2017 at 15:43

1 Answer 1

0
$\begingroup$

ANOVA is better suited for real-valued dependent variables, so in your case chi-square test is a better alternative. If you are doing this in R, you could use prop.test, it implements the same chi-squared when there are more than two groups (age groups, in your case).

$\endgroup$
5
  • $\begingroup$ Thanks a lot for your help, I appreciate it. I had also two other analysis, male-female success-unsuccess, and experience-not, success-unsuccess. Since those are tables 2 x 2 and with the kind of binary feature I had, I used chi square. Do you think that is OK, or you recommend something else? $\endgroup$ Aug 31, 2017 at 8:33
  • $\begingroup$ It's fine! There are other alternatives like tests for proportion, but chi-squared is perfectly valid as well. $\endgroup$ Aug 31, 2017 at 9:39
  • $\begingroup$ Thanks a lot for your help! Last question if I may ask. Thanks again! I am trying now to compare the results I got as metrics. The 29 participants would do specific tasks on 3 different websites. I have the average time that it took to complete each task on each website and the target. Can I do ANOVA to check if there is a significant difference between target vs website 1, 2, and 3? Or there is a better tool to compare the TARGET vs the real time it took? Thanks again $\endgroup$ Aug 31, 2017 at 10:05
  • $\begingroup$ Most likely ANOVA would do, but you have to be careful choosing between-subjects or within-subjects ANOVA. Do you have the same 29 participants doing the task on each of the 3 websites, or are they distributed between different websites? In the first case it is within-subjects ANOVA (or, as a nonparametric alternative, Friedman rank sum test); in the second ANOVA should be between-subjects (or you could use Kruskal-Wallis test). $\endgroup$ Aug 31, 2017 at 15:06
  • $\begingroup$ Thanks so much! I thought it would be ANOVA and now that I have your confirmation I will get to it. It is indeed 29 people doing all the 6 tasks; i didn't divide them. So I will use the within subject ANOVA. In one column I will put the target time and on the other one all the actual times. Does it sound right? Thank you!!!!! $\endgroup$ Aug 31, 2017 at 16:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.